The invention relates to the field of
machine learning of an
artificial intelligence technology, in particular to a method and a
system for identifying raw materials in a
kiln, which are mainly used for improving the
identification rate of raw materials in the
kiln and identifying risks in advance. The method comprises the following steps that S1, a
feature model M of the
raw material running in the
kiln is determined, a plurality of sets of data corresponding to the
feature model in off-line data in the kiln are collected to serve as samples, and the
feature model M comprises a plurality of feature parameters and data ranges corresponding to the feature parameters; s2, establishing a
multilayer perceptron (MLP) model for the feature model M; s3, training and testing a
multilayer perceptron (MLP) model according to the sample to obtain an optimal training model; and S4, detecting data corresponding to the characteristic parameters in the kiln condition in real time, inputting the data into the optimal training model, and determining the
raw material running condition according to the output of the optimal training model. The method solves the influence of human subjective factors, reduces the consumption of manpower, and improves the judgment rate.